研究背景：精神分裂症患者出院后得不到后续妥善的医疗照护，导致患者反复复发再入院。如何做好医院外的延伸护理是目前我国精神分裂症患者管理的重点。国外已有研究对慢性心力衰竭、糖尿病和高血压等患者进行分类，为后续制定针对性的干预措施提供依据。精神分裂症患者的延伸护理是否可以借鉴这种将患者分类后再针对性干预的方法，值得探讨。

研究目的：（1）根据精神分裂症患者的自我管理能力、自我效能及认知功能将患者进行分类，分析不同类型患者的特点；（2）以上述三个变量作为细分变量，构建精神分裂症患者亚型的判别模型。

研究方法：本研究采用邹海欧等发展的精神分裂症患者自我管理量表和重复成套神经心理状态评估工具（RBANS）对200例即将出院的精神分裂症患者进行问卷测评，通过SPSS 20.0中的聚类和判别分析方法对患者分类，并构建精神分裂症患者的亚型。

研究结果：研究样本中的200例即将出院的精神分裂症患者可分为5种类型：高认知自信自主型（26.5%）、高认知盲目自信型（20.5%）、高认知不自信不参与型（13.5%）、低认知盲目自信型（27.0%）和低认知不自信不参与型（12.5%）。患者的教育程度（P=0.002）、过去半年内服药状况（P <0.001）及主要照顾者的教育程度（P=0.027）在不同类型之间有显著差别。

研究结论：即将出院的精神分裂症患者存在不同类型，如医护人员有效甄别，研制针对不同类型患者的延伸护理干预，可更大程度地合理分配精神卫生资源并促进精神分裂症患者的康复。

Background: People with schizophrenia lack care after discharged from hospital which leads to relapse and recurrent readmissions. So, how to provide transitional care for people with schizophrenia is current focus. Some foreign studies classified people with chronic heart failure, diabetes and hypertension into different subtypes in order to provide targeted interventions. Whether transitional care for people with schizophrenia can draw on this method is worth researching.

Objectives: (1) To explore possible subtypes among people with schizophrenia before discharge by their self-management ability, self-efficacy and cognitive function status, then analyze characteristics of people with schizophrenia among different subtypes; (2) To build discrimination model for people with schizophrenia based on the above three variables.

Methods: Totally 200 Chinese people with schizophrenia before discharged from a tertiary psychiatric hospital in Beijing were assessed by Self-management Instrument for People with Schizophrenia and Repeatable Battery for the Assessment of Neuropsychological Status (RBANS). Cluster and discrimination analysis using SPSS20.0 package were performed to categorize and discriminate subjects based on their scores．

Results: Five different types of subjects were revealed in this sample: high cognition with confidence and autonomy (26.5%), high cognition with blind confidence (20.5%), high cognition with no participation (13.5%), low cognition with blind confidence (27.0%) and low cognition with no participation (12.5%). These five types of subjects had significant differences in patient's educational level (P = 0.002), use of antipsychotics in the past half year (P <0.001) and primary caregivers’ educational level (P = 0.027).

Conclusions: There are different subtypes among people with schizophrenia indeed. The finding may help health professionals give effective screening and targeted discharge measures which can further promote allocation of resources for mental health to a greater extent and promote the rehabilitation of people with schizophrenia.